Despite the importance and usefulness of basicity in metallurgical slags, the optimum quantitative expression of it has remained elusive, partly due to the fact that the precise concept of basicity has not been properly clarified. The two basic approaches to the concept of basicity are the Lux-Flood, which perceives basicity in terms of the activity of the free oxide ion, aO2-, and the Lewis, which perceives basicity in terms of the ability to donate negative charge. The former has been almost universally adopted in the area of metallurgical slags, while the latter has been increasingly accepted in the field of glass chemistry. In this paper, some difficulties of the Lux-Flood approach are pointed out, including the need for indirect measures of aO2-, which have proliferated in recent years. Accordingly, it is suggested that the Lewis approach has a good deal to commend it, and that the quantification of it in terms of optical basicity is the best measure of basicity available, which should therefore be more widely applied in metallurgical slags than it currently is. The fact that there are intrinsic problems in evaluating the optical basicities for the transition metal oxides is acknowledged, and a new scale is recommended. The patterns of iso-optical basicity lines in ternary systems are shown to agree well with the pattern of iso-activity and iso-capacity lines under appropriate conditions.